Realization of exponential filters and their application to biology/biomedicine
Loading...
Files
Date
2026-02-11
Authors
Λίτσο, Τζουλιάνο
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
The subject of this M.Sc. Thesis is the design and implementation of exponential filters based on Gaussian and Mittag–Leffler functions, with application to electrocardiogram (ECG) signal processing. These filters offer flexible frequency responses, which can be adjusted through the parameters of the Mittag–Leffler function, allowing better adaptation to the characteristics of biomedical signals.
The first part of this work focuses on the analysis of the ideal filter responses and their approximation in the frequency-domain using MATLAB. Since the ideal transfer functions cannot be directly implemented, a curve-fitting based method is applied in order to obtain rational integer-order transfer functions that closely follow the desired responses.
The obtained transfer functions are then implemented using analog circuit techniques. A fourth-order Inverse Follow-the-Leader Feedback (IFLF) topology based on Operational Transconductance Amplifiers (OTAs) is used to realize the filters, offering a systematic way to implement different filter responses without changing the basic structure. The performance of the proposed circuits is evaluated through schematic-level and post-layout simulations using Cadence tools and CMOS technology models.
In addition, the filters are implemented on a Field Programmable Analog Array (FPAA) platform, which allows experimental validation of the proposed designs. The experimental results confirm the effectiveness of the filters in reducing noise while preserving the main features of ECG signals.
Description
Keywords
Exponential filters, Gaussian filters, Mittag–Leffler filters